Codebook for processed data

Add data labels, metadata, codebook, distributions, and convert to other file types

Codebook

# Following codebook package's [vignette](https://cran.r-project.org/web/packages/codebook/vignettes/codebook_tutorial.html)

# Dependencies

library(tidyverse)
library(knitr)
library(kableExtra)
library(codebook)
library(labelled)
library(rio)

options(knitr.kable.NA = "/")


# Add labels ----

# read data
codebook_data <- read.csv("processed/4_data_participant_level_with_hand_scoring.csv")

# read data dictionary
dict <- read.csv("processed/5_data_dictionary.csv")

# add dictionary as labels
var_label(codebook_data) <- dict %>% 
  select(variable, label) %>% 
  dict_to_list()


# Add meta data ----

metadata(codebook_data)$name <- "Evaluative learning via deepfaked media"
metadata(codebook_data)$description <- "Across multiple experiments, we demonstrated that 'deepfakes' can establish automatic biases, self-reported evaluations, and behavioural intentions."
#metadata(codebook_data)$identifier <- "https://dx.doi.org/XXXXXXX"

metadata(codebook_data)$creator <- "Sean Hughes"
metadata(codebook_data)$citation <- "Hughes, S., Fried, O., Ferguson, M. J., Yao, D., Hughes, C., Hughes, R., & Hussey, I. (2020). Using Deepfakes to Hack the Human Mind."
metadata(codebook_data)$url <- "https://github.com/Sean-Hughes/DF-Impression-Formation--Video-and-Audio-"

# other meta data:  see https://schema.org/Dataset
metadata(codebook_data)$datePublished <- "2020"
metadata(codebook_data)$spatialCoverage <- "Online" 


# Create codebook ----

codebook(codebook_data)

Metadata

Description

Dataset name: Evaluative learning via deepfaked media

Across multiple experiments, we demonstrated that ‘deepfakes’ can establish automatic biases, self-reported evaluations, and behavioural intentions.

Metadata for search engines

• Spatial Coverage: Online

• Citation: Hughes, S., Fried, O., Ferguson, M. J., Yao, D., Hughes, C., Hughes, R., & Hussey, I. (2020). Using Deepfakes to Hack the Human Mind.

• URL: https://github.com/Sean-Hughes/DF-Impression-Formation–Video-and-Audio-

• Date published: 2020

• Creator:

  name	value
  1	Sean Hughes

x subject experiment intervention_medium source_valence experiment_condition exclude_subject exclude_implausible_intervention_linger intervention_linger_minutes age gender IAT_D2 mean_self_reported_evaluation mean_intentions deepfake_detected diagnosticity demand reactance hypothesis_awareness influence_awareness aot_actively_openminded_thinking_sum bcti_belief_in_conspiracy_sum crt_analytic_thinking_sum ocq_overclaiming_sum ras_relgious_affliation_sum rei_rational_sum rei_experiential_sum me_fake_news_awareness_sum me_real_news_awareness_sum me_fake_news_accuracy_sum me_real_news_accuracy_sum me_fake_news_sharing_sum me_real_news_sharing_sum aot_actively_openminded_thinking_pomp bcti_belief_in_conspiracy_pomp crt_analytic_thinking_pomp ocq_overclaiming_pomp ras_relgious_affliation_pomp rei_rational_pomp rei_experiential_pomp me_fake_news_awareness_pomp me_real_news_awareness_pomp me_fake_news_accuracy_pomp me_real_news_accuracy_pomp me_fake_news_sharing_pomp me_real_news_sharing_pomp IAT_D2_recoded_for_source_valence mean_self_reported_evaluation_recoded_for_source_valence mean_intentions_recoded_for_source_valence deepfake_concept_check deepfake_detected_rater_1 deepfake_detected_rater_2 deepfake_concept_check_rater_1 deepfake_concept_check_rater_2 gender_self_describe ethnicity ethnicity_self_describe location education employment education_recoded income income_recoded political_ideology_identity political_ideology_social_issues political_ideology_economic_issues religious_affiliation_general religious_affiliation_general_recoded passed_iat_performance complete_iat complete_selfreport complete_intentions task_order iat_block_order

#Variables

subject

Unique subject identifier

Distribution

Distribution of values for subject

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
subject	Unique subject identifier	character	0	1	1730	0	36	36	0

experiment

Experiment data was collected as part of

Distribution

Distribution of values for experiment

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
experiment	Experiment data was collected as part of	numeric	0	1	1	4	6	3.739306	1.482387	▆▇▇▅▅

intervention_medium

What medium did the intervention (whether deepfaked or genuine) take (e.g. video and audio vs just audio)

Distribution

Distribution of values for intervention_medium

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
intervention_medium	What medium did the intervention (whether deepfaked or genuine) take (e.g. video and audio vs just audio)	character	0	1	2	0	5	5	0

source_valence

What was the researcher-intended valence of the intervention? I.e. was the participant exposed to positive or negative messages?

Distribution

Distribution of values for source_valence

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
source_valence	What was the researcher-intended valence of the intervention? I.e. was the participant exposed to positive or negative messages?	character	0	1	2	0	8	8	0

experiment_condition

Was the intervention genuine or deepfaked content?

Distribution

Distribution of values for experiment_condition

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
experiment_condition	Was the intervention genuine or deepfaked content?	character	0	1	2	0	7	9	0

exclude_subject

Should the participant be excluded (TRUE) or retained (FALSE) based on passed_iat_performance complete_iat complete_selfreport and complete_intentions

Distribution

Distribution of values for exclude_subject

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
exclude_subject	Should the participant be excluded (TRUE) or retained (FALSE) based on passed_iat_performance complete_iat complete_selfreport and complete_intentions	logical	0	1	FAL: 1421, TRU: 309	0.1786127

exclude_implausible_intervention_linger

Should the participant be excluded (TRUE) or retained (FALSE) based on them spending not enough time (<1.5 minutes or too much time (>4.5 minutes) on the page that delivered the audio/video intervention.

Distribution

Distribution of values for exclude_implausible_intervention_linger

55 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
exclude_implausible_intervention_linger	Should the participant be excluded (TRUE) or retained (FALSE) based on them spending not enough time (<1.5 minutes or too much time (>4.5 minutes) on the page that delivered the audio/video intervention.	logical	55	0.9682081	TRU: 1555, FAL: 120	0.9283582

intervention_linger_minutes

Number of minutes spent viewing the page that delivered the intervention.

Distribution

Distribution of values for intervention_linger_minutes

55 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
intervention_linger_minutes	Number of minutes spent viewing the page that delivered the intervention.	numeric	55	0.9682081	0.028	2.7	2.7e+07	31805.05	920066.9	▇▁▁▁▁

age

Participant age

Distribution

Distribution of values for age

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
age	Participant age	numeric	0	1	18	29	70	31.24335	9.94172	▇▆▃▁▁

gender

Participant gender

Distribution

Distribution of values for gender

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
gender	Participant gender	character	0	1	5	0	4	22	0

IAT_D2

The IAT captures automatic evaluations of Chris the character depicted in the intervention. Accuracies and reaction times on IAT were converted to D2 scores following Greenwald et al 2003 and implemented using the IATScores R package. Briefly D2 is a standardized difference score of trimmed reaction times in one block type versus the other: (mean_block_2 - mean_block_1) / standard_deviation_all_blocks

Distribution

Distribution of values for IAT_D2

141 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
IAT_D2	The IAT captures automatic evaluations of Chris the character depicted in the intervention. Accuracies and reaction times on IAT were converted to D2 scores following Greenwald et al 2003 and implemented using the IATScores R package. Briefly D2 is a standardized difference score of trimmed reaction times in one block type versus the other: (mean_block_2 - mean_block_1) / standard_deviation_all_blocks	numeric	141	0.9184971	-1.2	0.23	1.6	0.2084644	0.3642037	▁▃▇▃▁

mean_self_reported_evaluation

Mean self-reported evaluation (positive-negative good-bad pleasant-unpleasant) of Chris the character depicted in the intervention.

Distribution

Distribution of values for mean_self_reported_evaluation

98 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
mean_self_reported_evaluation	Mean self-reported evaluation (positive-negative good-bad pleasant-unpleasant) of Chris the character depicted in the intervention.	numeric	98	0.9433526	-3	0	3	-0.0859804	2.029025	▇▃▃▅▇

mean_intentions

Mean behavioural intentions

Distribution

Distribution of values for mean_intentions

1485 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
mean_intentions	Mean behavioural intentions	numeric	1485	0.1416185	-2	-1	2	-0.9239184	1.052152	▇▅▃▂▁

deepfake_detected

Open ended responses to the deepfaking detection question were hand scored by two independent reviewers (see “3 Instructions for raters.docx”). If both reviewers scored an open ended response as having detected the intervention as a deepfake (whether correctly or not), this variable was set to TRUE. Otherwise FALSE.

Distribution

Distribution of values for deepfake_detected

1120 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
deepfake_detected	Open ended responses to the deepfaking detection question were hand scored by two independent reviewers (see “3 Instructions for raters.docx”). If both reviewers scored an open ended response as having detected the intervention as a deepfake (whether correctly or not), this variable was set to TRUE. Otherwise FALSE.	logical	1120	0.3526012	FAL: 493, TRU: 117	0.1918033

diagnosticity

Question assessing whether people believe that the targets actions in the video/audio were representative of his ‘true’ character

Distribution

Distribution of values for diagnosticity

70 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
diagnosticity	Question assessing whether people believe that the targets actions in the video/audio were representative of his ‘true’ character	numeric	70	0.9595376	0	2	3	1.966867	0.7643258	▁▃▁▇▃

demand

Assesssment of whether evaluations of the targer reflected how the person genuinely felt or if they simply responded in a way they thought the researchers wanted them to.

Distribution

Distribution of values for demand

131 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
demand	Assesssment of whether evaluations of the targer reflected how the person genuinely felt or if they simply responded in a way they thought the researchers wanted them to.	logical	131	0.9242775	FAL: 1585, TRU: 14	0.0087555

reactance

Assesssment of whether evaluations of the targer reflected how the person genuinely felt or if they responded in the opposite way than they thought the researchers wanted them to.

Distribution

Distribution of values for reactance

131 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
reactance	Assesssment of whether evaluations of the targer reflected how the person genuinely felt or if they responded in the opposite way than they thought the researchers wanted them to.	logical	131	0.9242775	FAL: 1537, TRU: 62	0.0387742

hypothesis_awareness

Question asking what participants thought the experiment was about (i.e. what the experimenter’s agenda was in the study)

Distribution

Distribution of values for hypothesis_awareness

1672 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
hypothesis_awareness	Question asking what participants thought the experiment was about (i.e. what the experimenter’s agenda was in the study)	logical	1672	0.033526	FAL: 58	0

influence_awareness

Question asking if participants thought the video/audio influenced how much they liked or disliked the target

Distribution

Distribution of values for influence_awareness

946 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
influence_awareness	Question asking if participants thought the video/audio influenced how much they liked or disliked the target	logical	946	0.4531792	TRU: 697, FAL: 87	0.8890306

aot_actively_openminded_thinking_sum

Scale assessing actively open minded thinking about evidence (Pennycook G. Cheyne J. A. Koehler D. & Fugelsang J. A. (2019). On the belief that beliefs should change according to evidence: Implications for conspiratorial moral paranormal political religious and science beliefs.)

Distribution

Distribution of values for aot_actively_openminded_thinking_sum

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
aot_actively_openminded_thinking_sum	Scale assessing actively open minded thinking about evidence (Pennycook G. Cheyne J. A. Koehler D. & Fugelsang J. A. (2019). On the belief that beliefs should change according to evidence: Implications for conspiratorial moral paranormal political religious and science beliefs.)	numeric	1486	0.1410405	22	38	48	37.40164	5.934706	▂▅▇▇▆

bcti_belief_in_conspiracy_sum

The Belief in Conspiracy Theories Inventory (BCTI, Swami et al., 2010) which assessed general conspiracist ideation or belief in conspiracy theories

Distribution

Distribution of values for bcti_belief_in_conspiracy_sum

1481 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
bcti_belief_in_conspiracy_sum	The Belief in Conspiracy Theories Inventory (BCTI, Swami et al., 2010) which assessed general conspiracist ideation or belief in conspiracy theories	numeric	1481	0.1439306	15	51	129	54.04819	23.60011	▇▇▆▂▁

crt_analytic_thinking_sum

The revised Cognitive Reflection Test (RCRT). The Revised Cognitive Reflection Test originally developed by Toplak West and Stanovich (2014) and subsequently revised by Bronstein Pennycook Bear Rand and Cannon (2019) was used to measure analytic thinking ability.

Distribution

Distribution of values for crt_analytic_thinking_sum

1236 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
crt_analytic_thinking_sum	The revised Cognitive Reflection Test (RCRT). The Revised Cognitive Reflection Test originally developed by Toplak West and Stanovich (2014) and subsequently revised by Bronstein Pennycook Bear Rand and Cannon (2019) was used to measure analytic thinking ability.	numeric	1236	0.2855491	0	4	7	3.953441	1.848155	▃▃▇▅▆

ocq_overclaiming_sum

The overclaiming questionnaire adapted from Paulhus et al. (2003).

Distribution

Distribution of values for ocq_overclaiming_sum

1480 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
ocq_overclaiming_sum	The overclaiming questionnaire adapted from Paulhus et al. (2003).	numeric	1480	0.1445087	10	72	143	71.516	29.66252	▃▇▇▆▂

ras_relgious_affliation_sum

The Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014).

Distribution

Distribution of values for ras_relgious_affliation_sum

1194 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
ras_relgious_affliation_sum	The Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014).	numeric	1194	0.3098266	8	21	40	20.77799	8.32119	▇▆▇▅▂

rei_rational_sum

The rational scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles.

Distribution

Distribution of values for rei_rational_sum

1235 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
rei_rational_sum	The rational scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles.	numeric	1235	0.2861272	15	52	70	50.79394	10.68324	▁▂▅▇▅

rei_experiential_sum

The experiential scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles.

Distribution

Distribution of values for rei_experiential_sum

1235 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
rei_experiential_sum	The experiential scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles.	numeric	1235	0.2861272	16	49	70	48.4101	8.768012	▁▂▇▇▂

me_fake_news_awareness_sum

Media evaluation task. Awareness sum score for the fake news items

Distribution

Distribution of values for me_fake_news_awareness_sum

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_fake_news_awareness_sum	Media evaluation task. Awareness sum score for the fake news items	numeric	1486	0.1410405	0	1	6	1.413934	1.191866	▇▃▂▁▁

me_real_news_awareness_sum

Media evaluation task. Awareness sum score for the real news items

Distribution

Distribution of values for me_real_news_awareness_sum

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_real_news_awareness_sum	Media evaluation task. Awareness sum score for the real news items	numeric	1486	0.1410405	0	0	4	0.2008197	0.5178296	▇▁▁▁▁

me_fake_news_accuracy_sum

Media evaluation task. Accuracy sum score for the fake news items

Distribution

Distribution of values for me_fake_news_accuracy_sum

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_fake_news_accuracy_sum	Media evaluation task. Accuracy sum score for the fake news items	numeric	1486	0.1410405	7	17	23	16.72131	2.614497	▁▂▇▇▂

me_real_news_accuracy_sum

Media evaluation task. Accuracy sum score for the real news items

Distribution

Distribution of values for me_real_news_accuracy_sum

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_real_news_accuracy_sum	Media evaluation task. Accuracy sum score for the real news items	numeric	1486	0.1410405	6	9	20	9.127049	2.460437	▇▆▂▁▁

me_fake_news_sharing_sum

Media evaluation task. Sharing sum score for the fake news items

Distribution

Distribution of values for me_fake_news_sharing_sum

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_fake_news_sharing_sum	Media evaluation task. Sharing sum score for the fake news items	numeric	1486	0.1410405	0	1	6	1.561475	1.768585	▇▂▁▂▁

me_real_news_sharing_sum

Media evaluation task. Accuracy sum score for the real news items

Distribution

Distribution of values for me_real_news_sharing_sum

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_real_news_sharing_sum	Media evaluation task. Accuracy sum score for the real news items	numeric	1486	0.1410405	0	0	6	0.454918	0.9739489	▇▁▁▁▁

aot_actively_openminded_thinking_pomp

POMP scores derived from scale assessing actively open minded thinking about evidence (Pennycook G. Cheyne J. A. Koehler D. & Fugelsang J. A. (2019). On the belief that beliefs should change according to evidence: Implications for conspiratorial moral paranormal political religious and science beliefs.)

Distribution

Distribution of values for aot_actively_openminded_thinking_pomp

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
aot_actively_openminded_thinking_pomp	POMP scores derived from scale assessing actively open minded thinking about evidence (Pennycook G. Cheyne J. A. Koehler D. & Fugelsang J. A. (2019). On the belief that beliefs should change according to evidence: Implications for conspiratorial moral paranormal political religious and science beliefs.)	numeric	1486	0.1410405	35	75	100	73.48361	14.87464	▂▅▇▇▆

bcti_belief_in_conspiracy_pomp

POMP scores derived from the Belief in Conspiracy Theories Inventory (BCTI, Swami et al., 2010) which assessed general conspiracist ideation or belief in conspiracy theories

Distribution

Distribution of values for bcti_belief_in_conspiracy_pomp

1481 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
bcti_belief_in_conspiracy_pomp	POMP scores derived from the Belief in Conspiracy Theories Inventory (BCTI, Swami et al., 2010) which assessed general conspiracist ideation or belief in conspiracy theories	numeric	1481	0.1439306	0	30	95	32.56627	19.67961	▇▇▆▂▁

crt_analytic_thinking_pomp

POMP scores derived from revised Cognitive Reflection Test (RCRT). The Revised Cognitive Reflection Test originally developed by Toplak West and Stanovich (2014) and subsequently revised by Bronstein Pennycook Bear Rand and Cannon (2019) was used to measure analytic thinking ability.

Distribution

Distribution of values for crt_analytic_thinking_pomp

1236 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
crt_analytic_thinking_pomp	POMP scores derived from revised Cognitive Reflection Test (RCRT). The Revised Cognitive Reflection Test originally developed by Toplak West and Stanovich (2014) and subsequently revised by Bronstein Pennycook Bear Rand and Cannon (2019) was used to measure analytic thinking ability.	numeric	1236	0.2855491	0	57	100	56.47368	26.39254	▃▃▇▅▆

ocq_overclaiming_pomp

POMP scores derived from the overclaiming questionnaire adapted from Paulhus et al. (2003).

Distribution

Distribution of values for ocq_overclaiming_pomp

1480 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
ocq_overclaiming_pomp	POMP scores derived from the overclaiming questionnaire adapted from Paulhus et al. (2003).	numeric	1480	0.1445087	6	40	79	39.728	16.45628	▃▇▇▆▂

ras_relgious_affliation_pomp

POMP scores derived from the Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014).

Distribution

Distribution of values for ras_relgious_affliation_pomp

1194 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
ras_relgious_affliation_pomp	POMP scores derived from the Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014).	numeric	1194	0.3098266	0	41	100	39.91978	26.0168	▇▆▇▅▂

rei_rational_pomp

POMP scores for the rational scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles.

Distribution

Distribution of values for rei_rational_pomp

1235 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
rei_rational_pomp	POMP scores for the rational scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles.	numeric	1235	0.2861272	-4	27	42	25.67879	8.894561	▁▂▅▇▅

rei_experiential_pomp

POMP scores for the experiential scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles.

Distribution

Distribution of values for rei_experiential_pomp

1235 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
rei_experiential_pomp	POMP scores for the experiential scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles.	numeric	1235	0.2861272	-3	24	42	23.68687	7.323976	▁▂▇▇▂

me_fake_news_awareness_pomp

Media evaluation task. POMP score for the fake news items (awareness)

Distribution

Distribution of values for me_fake_news_awareness_pomp

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_fake_news_awareness_pomp	Media evaluation task. POMP score for the fake news items (awareness)	numeric	1486	0.1410405	-50	-46	-25	-44.29508	4.888364	▇▃▂▁▁

me_real_news_awareness_pomp

Media evaluation task. POMP score for the real news items (awareness)

Distribution

Distribution of values for me_real_news_awareness_pomp

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_real_news_awareness_pomp	Media evaluation task. POMP score for the real news items (awareness)	numeric	1486	0.1410405	-50	-50	-33	-49.19262	2.102269	▇▁▁▁▁

me_fake_news_accuracy_pomp

Media evaluation task. POMP score for the fake news items (accuracy)

Distribution

Distribution of values for me_fake_news_accuracy_pomp

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_fake_news_accuracy_pomp	Media evaluation task. POMP score for the fake news items (accuracy)	numeric	1486	0.1410405	-14	14	31	13.11475	7.217602	▁▂▆▇▁

me_real_news_accuracy_pomp

Media evaluation task. POMP score for the real news items (accuracy)

Distribution

Distribution of values for me_real_news_accuracy_pomp

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_real_news_accuracy_pomp	Media evaluation task. POMP score for the real news items (accuracy)	numeric	1486	0.1410405	-17	-8	22	-8.032787	6.895379	▇▆▂▁▁

me_fake_news_sharing_pomp

Media evaluation task. POMP score for the fake news items (sharing)

Distribution

Distribution of values for me_fake_news_sharing_pomp

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_fake_news_sharing_pomp	Media evaluation task. POMP score for the fake news items (sharing)	numeric	1486	0.1410405	-50	-46	-25	-43.57787	7.389103	▇▂▁▂▁

me_real_news_sharing_pomp

Media evaluation task. POMP score for the real news items (sharing)

Distribution

Distribution of values for me_real_news_sharing_pomp

1486 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
me_real_news_sharing_pomp	Media evaluation task. POMP score for the real news items (sharing)	numeric	1486	0.1410405	-50	-50	-25	-48.15984	3.989062	▇▁▁▁▁

IAT_D2_recoded_for_source_valence

IAT_D2 recoded for source_valence. If source_valence == “negative”, IAT_D2*-1, otherwise IAT_D2.

Distribution

Distribution of values for IAT_D2_recoded_for_source_valence

141 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
IAT_D2_recoded_for_source_valence	IAT_D2 recoded for source_valence. If source_valence == “negative”, IAT_D2*-1, otherwise IAT_D2.	numeric	141	0.9184971	-0.94	0.2	1.6	0.1854097	0.3764714	▁▅▇▃▁

mean_self_reported_evaluation_recoded_for_source_valence

mean_self_reported_evaluation recoded for source_valence. If source_valence == “negative”, mean_self_reported_evaluation*-1, otherwise mean_self_reported_evaluation.

Distribution

Distribution of values for mean_self_reported_evaluation_recoded_for_source_valence

98 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
mean_self_reported_evaluation_recoded_for_source_valence	mean_self_reported_evaluation recoded for source_valence. If source_valence == “negative”, mean_self_reported_evaluation*-1, otherwise mean_self_reported_evaluation.	numeric	98	0.9433526	-3	2	3	1.568836	1.289025	▁▁▂▃▇

mean_intentions_recoded_for_source_valence

mean_intentions recoded for source_valence. If source_valence == “negative”, mean_intentions*-1, otherwise mean_intentions.

Distribution

Distribution of values for mean_intentions_recoded_for_source_valence

1485 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
mean_intentions_recoded_for_source_valence	mean_intentions recoded for source_valence. If source_valence == “negative”, mean_intentions*-1, otherwise mean_intentions.	numeric	1485	0.1416185	-2	0.67	2	0.5263673	1.298441	▂▃▅▂▇

deepfake_concept_check

Question assessin if participants were aware of the concept of Deepfaking before taking part in the study

Distribution

Distribution of values for deepfake_concept_check

1294 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
deepfake_concept_check	Question assessin if participants were aware of the concept of Deepfaking before taking part in the study	logical	1294	0.2520231	TRU: 234, FAL: 202	0.5366972

deepfake_detected_rater_1

Ratings indicating whether a participant detected that they were exposed to Deepfaked content during the study. Ratings obtained from Rater 1

Distribution

Distribution of values for deepfake_detected_rater_1

1120 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
deepfake_detected_rater_1	Ratings indicating whether a participant detected that they were exposed to Deepfaked content during the study. Ratings obtained from Rater 1	logical	1120	0.3526012	FAL: 462, TRU: 148	0.242623

deepfake_detected_rater_2

Ratings indicating whether a participant detected that they were exposed to Deepfaked content during the study. Ratings obtained from Rater 2

Distribution

Distribution of values for deepfake_detected_rater_2

1120 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
deepfake_detected_rater_2	Ratings indicating whether a participant detected that they were exposed to Deepfaked content during the study. Ratings obtained from Rater 2	logical	1120	0.3526012	FAL: 476, TRU: 134	0.2196721

deepfake_concept_check_rater_1

Ratings indicating if participants were aware of the concept of a Deepfake before taking part in the study. Ratings obtained from Rater 1

Distribution

Distribution of values for deepfake_concept_check_rater_1

1294 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
deepfake_concept_check_rater_1	Ratings indicating if participants were aware of the concept of a Deepfake before taking part in the study. Ratings obtained from Rater 1	logical	1294	0.2520231	TRU: 247, FAL: 189	0.5665138

deepfake_concept_check_rater_2

Ratings indicating if participants were aware of the concept of a Deepfake before taking part in the study. Ratings obtained from Rater 2

Distribution

Distribution of values for deepfake_concept_check_rater_2

1294 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
deepfake_concept_check_rater_2	Ratings indicating if participants were aware of the concept of a Deepfake before taking part in the study. Ratings obtained from Rater 2	logical	1294	0.2520231	TRU: 247, FAL: 189	0.5665138

gender_self_describe

Open ended gender

Distribution

Distribution of values for gender_self_describe

1725 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
gender_self_describe	Open ended gender	character	1725	0.0028902	1	0	4	4	0

ethnicity

Ethnicity

Distribution

Distribution of values for ethnicity

1189 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
ethnicity	Ethnicity	character	1189	0.3127168	7	0	5	25	0

ethnicity_self_describe

Open ended ethnicity

Distribution

Distribution of values for ethnicity_self_describe

1701 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
ethnicity_self_describe	Open ended ethnicity	character	1701	0.016763	19	0	4	40	0

location

Location

Distribution

Distribution of values for location

1190 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
location	Location	character	1190	0.3121387	18	0	5	24	0

education

Education

Distribution

Distribution of values for education

1189 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
education	Education	character	1189	0.3127168	8	0	7	21	0

employment

Employment status

Distribution

Distribution of values for employment

1189 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
employment	Employment status	character	1189	0.3127168	10	0	7	24	0

education_recoded

Education recoded into seven groups

Distribution

Distribution of values for education_recoded

1189 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
education_recoded	Education recoded into seven groups	numeric	1189	0.3127168	1	5	7	4.072089	1.526427	▅▅▁▇▃

income

Income level

Distribution

Distribution of values for income

1190 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
income	Income level	character	1190	0.3121387	10	0	12	20	0

income_recoded

Income level recoded into 7 groups

Distribution

Distribution of values for income_recoded

1238 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
income_recoded	Income level recoded into 7 groups	numeric	1238	0.2843931	1	3	8	2.943089	1.503835	▇▆▇▁▁

political_ideology_identity

4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about the importance of political ideology to one’s self-identity

Distribution

Distribution of values for political_ideology_identity

1190 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
political_ideology_identity	4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about the importance of political ideology to one’s self-identity	numeric	1190	0.3121387	-3	1	3	0.8574074	1.544648	▂▂▃▇▇

political_ideology_social_issues

4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about how liberal vs. conservative they are when it comes to social issues.

Distribution

Distribution of values for political_ideology_social_issues

1190 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
political_ideology_social_issues	4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about how liberal vs. conservative they are when it comes to social issues.	numeric	1190	0.3121387	-2	-1	2	-0.637037	1.023707	▅▇▆▂▁

political_ideology_economic_issues

4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about how liberal vs. conservative they are when it comes to economic issues.

Distribution

Distribution of values for political_ideology_economic_issues

1190 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	min	median	max	mean	sd	hist
political_ideology_economic_issues	4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about how liberal vs. conservative they are when it comes to economic issues.	numeric	1190	0.3121387	-2	0	2	-0.4148148	1.032929	▃▇▇▃▁

religious_affiliation_general

Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014). Raw responses

Distribution

Distribution of values for religious_affiliation_general

1190 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
religious_affiliation_general	Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014). Raw responses	character	1190	0.3121387	10	0	4	16	0

religious_affiliation_general_recoded

Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014). Participants categorised into one of three groups (Agnostic Atheist Religious)

Distribution

Distribution of values for religious_affiliation_general_recoded

1190 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
religious_affiliation_general_recoded	Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014). Participants categorised into one of three groups (Agnostic Atheist Religious)	character	1190	0.3121387	3	0	7	9	0

passed_iat_performance

Mark participants for exclusion if their total error rate is >30% their error rate in any one block is >40% or if >10% RTs are <300ms.

Distribution

Distribution of values for passed_iat_performance

122 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	count	mean
passed_iat_performance	Mark participants for exclusion if their total error rate is >30% their error rate in any one block is >40% or if >10% RTs are <300ms.	logical	122	0.9294798	TRU: 1440, FAL: 168	0.8955224

complete_iat

Complete IAT data (used for exclusions)

Distribution

Distribution of values for complete_iat

122 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
complete_iat	Complete IAT data (used for exclusions)	character	122	0.9294798	3	0	6	8	0

complete_selfreport

Complete self reported evaluations data (used for exclusions)

Distribution

Distribution of values for complete_selfreport

98 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
complete_selfreport	Complete self reported evaluations data (used for exclusions)	character	98	0.9433526	2	0	6	8	0

complete_intentions

Complete behavioural intentions data (used for exclusions)

Distribution

Distribution of values for complete_intentions

1485 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
complete_intentions	Complete behavioural intentions data (used for exclusions)	character	1485	0.1416185	2	0	6	8	0

task_order

Did the particiapant complete the self-reported evaluations or the IAT first?

Distribution

Distribution of values for task_order

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
task_order	Did the particiapant complete the self-reported evaluations or the IAT first?	character	0	1	2	0	9	25	0

iat_block_order

Was the participant exposed to the intervention-consistent or intervention-inconsistent block in the IAT first? E.g. if source_valence was negative did the IAT map Chris and negative to the same key (con) or Chris and positive (incon)?

Distribution

Distribution of values for iat_block_order

0 missing values.

Summary statistics

name	label	data_type	n_missing	complete_rate	n_unique	empty	min	max	whitespace
iat_block_order	Was the participant exposed to the intervention-consistent or intervention-inconsistent block in the IAT first? E.g. if source_valence was negative did the IAT map Chris and negative to the same key (con) or Chris and positive (incon)?	character	0	1	2	0	31	33	0

Missingness report

Codebook table

JSON-LD metadata

The following JSON-LD can be found by search engines, if you share this codebook publicly on the web.

{
  "name": "Evaluative learning via deepfaked media",
  "description": "Across multiple experiments, we demonstrated that 'deepfakes' can establish automatic biases, self-reported evaluations, and behavioural intentions.\n\n\n## Table of variables\nThis table contains variable names, labels, and number of missing values.\nSee the complete codebook for more.\n\n[truncated]\n\n### Note\nThis dataset was automatically described using the [codebook R package](https://rubenarslan.github.io/codebook/) (version 0.9.2).",
  "creator": "Sean Hughes",
  "citation": "Hughes, S., Fried, O., Ferguson, M. J., Yao, D., Hughes, C., Hughes, R., & Hussey, I. (2020). Using Deepfakes to Hack the Human Mind.",
  "url": "https://github.com/Sean-Hughes/DF-Impression-Formation--Video-and-Audio-",
  "datePublished": "2020",
  "spatialCoverage": "Online",
  "keywords": ["subject", "experiment", "intervention_medium", "source_valence", "experiment_condition", "exclude_subject", "exclude_implausible_intervention_linger", "intervention_linger_minutes", "age", "gender", "IAT_D2", "mean_self_reported_evaluation", "mean_intentions", "deepfake_detected", "diagnosticity", "demand", "reactance", "hypothesis_awareness", "influence_awareness", "aot_actively_openminded_thinking_sum", "bcti_belief_in_conspiracy_sum", "crt_analytic_thinking_sum", "ocq_overclaiming_sum", "ras_relgious_affliation_sum", "rei_rational_sum", "rei_experiential_sum", "me_fake_news_awareness_sum", "me_real_news_awareness_sum", "me_fake_news_accuracy_sum", "me_real_news_accuracy_sum", "me_fake_news_sharing_sum", "me_real_news_sharing_sum", "aot_actively_openminded_thinking_pomp", "bcti_belief_in_conspiracy_pomp", "crt_analytic_thinking_pomp", "ocq_overclaiming_pomp", "ras_relgious_affliation_pomp", "rei_rational_pomp", "rei_experiential_pomp", "me_fake_news_awareness_pomp", "me_real_news_awareness_pomp", "me_fake_news_accuracy_pomp", "me_real_news_accuracy_pomp", "me_fake_news_sharing_pomp", "me_real_news_sharing_pomp", "IAT_D2_recoded_for_source_valence", "mean_self_reported_evaluation_recoded_for_source_valence", "mean_intentions_recoded_for_source_valence", "deepfake_concept_check", "deepfake_detected_rater_1", "deepfake_detected_rater_2", "deepfake_concept_check_rater_1", "deepfake_concept_check_rater_2", "gender_self_describe", "ethnicity", "ethnicity_self_describe", "location", "education", "employment", "education_recoded", "income", "income_recoded", "political_ideology_identity", "political_ideology_social_issues", "political_ideology_economic_issues", "religious_affiliation_general", "religious_affiliation_general_recoded", "passed_iat_performance", "complete_iat", "complete_selfreport", "complete_intentions", "task_order", "iat_block_order"],
  "@context": "http://schema.org/",
  "@type": "Dataset",
  "variableMeasured": [
    {
      "name": "subject",
      "description": "Unique subject identifier",
      "@type": "propertyValue"
    },
    {
      "name": "experiment",
      "description": "Experiment data was collected as part of",
      "@type": "propertyValue"
    },
    {
      "name": "intervention_medium",
      "description": "What medium did the intervention (whether deepfaked or genuine) take (e.g. video and audio vs just audio)",
      "@type": "propertyValue"
    },
    {
      "name": "source_valence",
      "description": "What was the researcher-intended valence of the intervention? I.e. was the participant exposed to positive or negative messages?",
      "@type": "propertyValue"
    },
    {
      "name": "experiment_condition",
      "description": "Was the intervention genuine or deepfaked content?",
      "@type": "propertyValue"
    },
    {
      "name": "exclude_subject",
      "description": "Should the participant be excluded (TRUE) or retained (FALSE) based on passed_iat_performance complete_iat complete_selfreport and complete_intentions",
      "@type": "propertyValue"
    },
    {
      "name": "exclude_implausible_intervention_linger",
      "description": "Should the participant be excluded (TRUE) or retained (FALSE) based on them spending not enough time (<1.5 minutes or too much time (>4.5 minutes) on the page that delivered the audio/video intervention.",
      "@type": "propertyValue"
    },
    {
      "name": "intervention_linger_minutes",
      "description": "Number of minutes spent viewing the page that delivered the intervention.",
      "@type": "propertyValue"
    },
    {
      "name": "age",
      "description": "Participant age",
      "@type": "propertyValue"
    },
    {
      "name": "gender",
      "description": "Participant gender",
      "@type": "propertyValue"
    },
    {
      "name": "IAT_D2",
      "description": "The IAT captures automatic evaluations of Chris the character depicted in the intervention. Accuracies and reaction times on IAT were converted to D2 scores following Greenwald et al 2003 and implemented using the IATScores R package. Briefly D2 is a standardized difference score of trimmed reaction times in one block type versus the other: (mean_block_2 - mean_block_1) / standard_deviation_all_blocks",
      "@type": "propertyValue"
    },
    {
      "name": "mean_self_reported_evaluation",
      "description": "Mean self-reported evaluation (positive-negative good-bad pleasant-unpleasant) of Chris the character depicted in the intervention.",
      "@type": "propertyValue"
    },
    {
      "name": "mean_intentions",
      "description": "Mean behavioural intentions",
      "@type": "propertyValue"
    },
    {
      "name": "deepfake_detected",
      "description": "Open ended responses to the deepfaking detection question were hand scored by two independent reviewers (see \"3 Instructions for raters.docx\"). If both reviewers scored an open ended response as having detected the intervention as a deepfake (whether correctly or not), this variable was set to TRUE. Otherwise FALSE.",
      "@type": "propertyValue"
    },
    {
      "name": "diagnosticity",
      "description": "Question assessing whether people believe that the targets actions in the video/audio were representative of his 'true' character",
      "@type": "propertyValue"
    },
    {
      "name": "demand",
      "description": "Assesssment of whether evaluations of the targer reflected how the person genuinely felt or if they simply responded in a way they thought the researchers wanted them to.",
      "@type": "propertyValue"
    },
    {
      "name": "reactance",
      "description": "Assesssment of whether evaluations of the targer reflected how the person genuinely felt or if they responded in the opposite way than they thought the researchers wanted them to.",
      "@type": "propertyValue"
    },
    {
      "name": "hypothesis_awareness",
      "description": "Question asking what participants thought the experiment was about (i.e. what the experimenter's agenda was in the study)",
      "@type": "propertyValue"
    },
    {
      "name": "influence_awareness",
      "description": "Question asking if participants thought the video/audio influenced how much they liked or disliked the target",
      "@type": "propertyValue"
    },
    {
      "name": "aot_actively_openminded_thinking_sum",
      "description": "Scale assessing actively open minded thinking about evidence (Pennycook G. Cheyne J. A. Koehler D. & Fugelsang J. A. (2019). On the belief that beliefs should change according to evidence: Implications for conspiratorial moral paranormal political religious and science beliefs.)",
      "@type": "propertyValue"
    },
    {
      "name": "bcti_belief_in_conspiracy_sum",
      "description": "The Belief in Conspiracy Theories Inventory (BCTI, Swami et al., 2010) which assessed general conspiracist ideation or belief in conspiracy theories ",
      "@type": "propertyValue"
    },
    {
      "name": "crt_analytic_thinking_sum",
      "description": "The revised Cognitive Reflection Test (RCRT). The Revised Cognitive Reflection Test originally developed by Toplak West and Stanovich (2014) and subsequently revised by Bronstein Pennycook Bear Rand and Cannon (2019) was used to measure analytic thinking ability. ",
      "@type": "propertyValue"
    },
    {
      "name": "ocq_overclaiming_sum",
      "description": "The overclaiming questionnaire adapted from Paulhus et al. (2003). ",
      "@type": "propertyValue"
    },
    {
      "name": "ras_relgious_affliation_sum",
      "description": "The Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014). ",
      "@type": "propertyValue"
    },
    {
      "name": "rei_rational_sum",
      "description": "The rational scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles. ",
      "@type": "propertyValue"
    },
    {
      "name": "rei_experiential_sum",
      "description": "The experiential scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles. ",
      "@type": "propertyValue"
    },
    {
      "name": "me_fake_news_awareness_sum",
      "description": "Media evaluation task. Awareness sum score for the fake news items",
      "@type": "propertyValue"
    },
    {
      "name": "me_real_news_awareness_sum",
      "description": "Media evaluation task. Awareness sum score for the real news items",
      "@type": "propertyValue"
    },
    {
      "name": "me_fake_news_accuracy_sum",
      "description": "Media evaluation task. Accuracy sum score for the fake news items",
      "@type": "propertyValue"
    },
    {
      "name": "me_real_news_accuracy_sum",
      "description": "Media evaluation task. Accuracy sum score for the real news items",
      "@type": "propertyValue"
    },
    {
      "name": "me_fake_news_sharing_sum",
      "description": "Media evaluation task. Sharing sum score for the fake news items",
      "@type": "propertyValue"
    },
    {
      "name": "me_real_news_sharing_sum",
      "description": "Media evaluation task. Accuracy sum score for the real news items",
      "@type": "propertyValue"
    },
    {
      "name": "aot_actively_openminded_thinking_pomp",
      "description": "POMP scores derived from scale assessing actively open minded thinking about evidence (Pennycook G. Cheyne J. A. Koehler D. & Fugelsang J. A. (2019). On the belief that beliefs should change according to evidence: Implications for conspiratorial moral paranormal political religious and science beliefs.)",
      "@type": "propertyValue"
    },
    {
      "name": "bcti_belief_in_conspiracy_pomp",
      "description": "POMP scores derived from the Belief in Conspiracy Theories Inventory (BCTI, Swami et al., 2010) which assessed general conspiracist ideation or belief in conspiracy theories ",
      "@type": "propertyValue"
    },
    {
      "name": "crt_analytic_thinking_pomp",
      "description": "POMP scores derived from revised Cognitive Reflection Test (RCRT). The Revised Cognitive Reflection Test originally developed by Toplak West and Stanovich (2014) and subsequently revised by Bronstein Pennycook Bear Rand and Cannon (2019) was used to measure analytic thinking ability. ",
      "@type": "propertyValue"
    },
    {
      "name": "ocq_overclaiming_pomp",
      "description": "POMP scores derived from the overclaiming questionnaire adapted from Paulhus et al. (2003). ",
      "@type": "propertyValue"
    },
    {
      "name": "ras_relgious_affliation_pomp",
      "description": "POMP scores derived from the Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014). ",
      "@type": "propertyValue"
    },
    {
      "name": "rei_rational_pomp",
      "description": "POMP scores for the rational scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles. ",
      "@type": "propertyValue"
    },
    {
      "name": "rei_experiential_pomp",
      "description": "POMP scores for the experiential scale. Derived from the Rational-Experiential Inventory (REI) developed by Pacini and Epstein (1999) designed to measure individual differences in processing styles. ",
      "@type": "propertyValue"
    },
    {
      "name": "me_fake_news_awareness_pomp",
      "description": "Media evaluation task. POMP score for the fake news items (awareness)",
      "@type": "propertyValue"
    },
    {
      "name": "me_real_news_awareness_pomp",
      "description": "Media evaluation task. POMP score for the real news items (awareness)",
      "@type": "propertyValue"
    },
    {
      "name": "me_fake_news_accuracy_pomp",
      "description": "Media evaluation task. POMP score for the fake news items (accuracy)",
      "@type": "propertyValue"
    },
    {
      "name": "me_real_news_accuracy_pomp",
      "description": "Media evaluation task. POMP score for the real news items (accuracy)",
      "@type": "propertyValue"
    },
    {
      "name": "me_fake_news_sharing_pomp",
      "description": "Media evaluation task. POMP score for the fake news items (sharing)",
      "@type": "propertyValue"
    },
    {
      "name": "me_real_news_sharing_pomp",
      "description": "Media evaluation task. POMP score for the real news items (sharing)",
      "@type": "propertyValue"
    },
    {
      "name": "IAT_D2_recoded_for_source_valence",
      "description": "IAT_D2 recoded for source_valence. If source_valence == \"negative\", IAT_D2*-1, otherwise IAT_D2.",
      "@type": "propertyValue"
    },
    {
      "name": "mean_self_reported_evaluation_recoded_for_source_valence",
      "description": "mean_self_reported_evaluation recoded for source_valence. If source_valence == \"negative\", mean_self_reported_evaluation*-1, otherwise mean_self_reported_evaluation.",
      "@type": "propertyValue"
    },
    {
      "name": "mean_intentions_recoded_for_source_valence",
      "description": "mean_intentions recoded for source_valence. If source_valence == \"negative\", mean_intentions*-1, otherwise mean_intentions.",
      "@type": "propertyValue"
    },
    {
      "name": "deepfake_concept_check",
      "description": "Question assessin if participants were aware of the concept of Deepfaking before taking part in the study",
      "@type": "propertyValue"
    },
    {
      "name": "deepfake_detected_rater_1",
      "description": "Ratings indicating whether a participant detected that they were exposed to Deepfaked content during the study. Ratings obtained from Rater 1",
      "@type": "propertyValue"
    },
    {
      "name": "deepfake_detected_rater_2",
      "description": "Ratings indicating whether a participant detected that they were exposed to Deepfaked content during the study. Ratings obtained from Rater 2",
      "@type": "propertyValue"
    },
    {
      "name": "deepfake_concept_check_rater_1",
      "description": "Ratings indicating if participants were aware of the concept of a Deepfake before  taking part in the study. Ratings obtained from Rater 1",
      "@type": "propertyValue"
    },
    {
      "name": "deepfake_concept_check_rater_2",
      "description": "Ratings indicating if participants were aware of the concept of a Deepfake before  taking part in the study. Ratings obtained from Rater 2",
      "@type": "propertyValue"
    },
    {
      "name": "gender_self_describe",
      "description": "Open ended gender",
      "@type": "propertyValue"
    },
    {
      "name": "ethnicity",
      "description": "Ethnicity",
      "@type": "propertyValue"
    },
    {
      "name": "ethnicity_self_describe",
      "description": "Open ended ethnicity",
      "@type": "propertyValue"
    },
    {
      "name": "location",
      "description": "Location",
      "@type": "propertyValue"
    },
    {
      "name": "education",
      "description": "Education",
      "@type": "propertyValue"
    },
    {
      "name": "employment",
      "description": "Employment status",
      "@type": "propertyValue"
    },
    {
      "name": "education_recoded",
      "description": "Education recoded into seven groups",
      "@type": "propertyValue"
    },
    {
      "name": "income",
      "description": "Income level",
      "@type": "propertyValue"
    },
    {
      "name": "income_recoded",
      "description": "Income level recoded into 7 groups",
      "@type": "propertyValue"
    },
    {
      "name": "political_ideology_identity",
      "description": "4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about the importance of political ideology to one's self-identity",
      "@type": "propertyValue"
    },
    {
      "name": "political_ideology_social_issues",
      "description": "4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about how liberal vs. conservative they are when it comes to social issues.",
      "@type": "propertyValue"
    },
    {
      "name": "political_ideology_economic_issues",
      "description": "4 item-measure of political ideology developed by Pennycook and Rand (2018). This measure combines the two questions about how liberal vs. conservative they are when it comes to economic issues.",
      "@type": "propertyValue"
    },
    {
      "name": "religious_affiliation_general",
      "description": "Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014). Raw responses",
      "@type": "propertyValue"
    },
    {
      "name": "religious_affiliation_general_recoded",
      "description": "Religious Affiliation Scale (Pennycook Cheyne Barr Koehler & Fugelsang 2014). Participants categorised into one of three groups (Agnostic Atheist Religious)",
      "@type": "propertyValue"
    },
    {
      "name": "passed_iat_performance",
      "description": "Mark participants for exclusion if their total error rate is >30% their error rate in any one block is >40% or if >10% RTs are <300ms.",
      "@type": "propertyValue"
    },
    {
      "name": "complete_iat",
      "description": "Complete IAT data (used for exclusions)",
      "@type": "propertyValue"
    },
    {
      "name": "complete_selfreport",
      "description": "Complete self reported evaluations data (used for exclusions)",
      "@type": "propertyValue"
    },
    {
      "name": "complete_intentions",
      "description": "Complete behavioural intentions data (used for exclusions)",
      "@type": "propertyValue"
    },
    {
      "name": "task_order",
      "description": "Did the particiapant complete the self-reported evaluations or the IAT first?",
      "@type": "propertyValue"
    },
    {
      "name": "iat_block_order",
      "description": "Was the participant exposed to the intervention-consistent or intervention-inconsistent block in the IAT first? E.g. if source_valence was negative did the IAT map Chris and negative to the same key (con) or Chris and positive (incon)?",
      "@type": "propertyValue"
    }
  ]
}`

Write alternative file types to disk

Original csv file is used for analyses (as it is simplest), but other file types that integrate the labels are likely to be more useful for reuse.

I include an R .rds file (which includes data labels and data types), SPSS .sav and Stata .dta files.

write_rds(codebook_data, "processed/4_data_participant_level_with_hand_scoring.rds") 
rio::export(codebook_data, "processed/4_data_participant_level_with_hand_scoring.sav") # SPSS file
rio::export(codebook_data, "processed/4_data_participant_level_with_hand_scoring.dta") # Stata file

Session Info

sessionInfo()

## R version 4.0.2 (2020-06-22)
## Platform: x86_64-apple-darwin17.0 (64-bit)
## Running under: macOS Catalina 10.15.7
## 
## Matrix products: default
## BLAS:   /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRblas.dylib
## LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib
## 
## locale:
## [1] en_IE.UTF-8/en_IE.UTF-8/en_IE.UTF-8/C/en_IE.UTF-8/en_IE.UTF-8
## 
## attached base packages:
## [1] stats     graphics  grDevices utils     datasets  methods   base     
## 
## other attached packages:
##  [1] rio_0.5.16       labelled_2.7.0   codebook_0.9.2   kableExtra_1.3.1
##  [5] knitr_1.30       forcats_0.5.0    stringr_1.4.0    dplyr_1.0.2     
##  [9] purrr_0.3.4      readr_1.3.1      tidyr_1.1.2      tibble_3.0.3    
## [13] ggplot2_3.3.2    tidyverse_1.3.0 
## 
## loaded via a namespace (and not attached):
##  [1] httr_1.4.1        jsonlite_1.7.1    viridisLite_0.3.0 modelr_0.1.8     
##  [5] assertthat_0.2.1  highr_0.8         cellranger_1.1.0  yaml_2.2.1       
##  [9] globals_0.13.1    pillar_1.4.6      backports_1.1.9   glue_1.4.2       
## [13] digest_0.6.25     rvest_0.3.5       colorspace_1.4-1  htmltools_0.5.0  
## [17] pkgconfig_2.0.3   broom_0.7.2       listenv_0.8.0     haven_2.3.1      
## [21] scales_1.1.1      webshot_0.5.2     openxlsx_4.1.5    generics_0.0.2   
## [25] farver_2.0.3      ellipsis_0.3.1    DT_0.13           withr_2.2.0      
## [29] repr_1.1.0        skimr_2.1.2       cli_2.0.2         rmdpartials_0.5.8
## [33] magrittr_1.5      crayon_1.3.4      readxl_1.3.1      evaluate_0.14    
## [37] fs_1.4.1          future_1.19.1     fansi_0.4.1       xml2_1.3.2       
## [41] foreign_0.8-80    tools_4.0.2       data.table_1.13.2 hms_0.5.3        
## [45] lifecycle_0.2.0   munsell_0.5.0     reprex_0.3.0      zip_2.1.1        
## [49] compiler_4.0.2    rlang_0.4.8       grid_4.0.2        rstudioapi_0.11  
## [53] htmlwidgets_1.5.1 crosstalk_1.1.0.1 base64enc_0.1-3   labeling_0.3     
## [57] rmarkdown_2.5     gtable_0.3.0      codetools_0.2-16  DBI_1.1.0        
## [61] curl_4.3          R6_2.4.1          lubridate_1.7.9   stringi_1.4.6    
## [65] parallel_4.0.2    Rcpp_1.0.5        vctrs_0.3.4       dbplyr_1.4.3     
## [69] tidyselect_1.1.0  xfun_0.19